One-pot synthesis of terpyridines and macrocyclization to C3-symmetric cyclosexipyridines

Article abstract of DOI:10.1139/v03-020

Four examples of 2,6-dicinnamoylpyridines were obtained in 60-65% yields in condensations of commercially available 2,6-diacetylpyridine and benzaldehydes in 1:2 stoichiometry. At 2:1 ratios, four related 6,6″-diacetylated-4′-arylterpyridines were isolate

Full text of DOI:10.1139/v03-020

209
One-pot synthesis of terpyridines and
macrocyclization to C3-symmetric
cyclosexipyridines
Lucie Masciello and Pierre G. Potvin
Abstract: Four examples of 2,6-dicinnamoylpyridines were obtained in 60–65% yields in condensations of commer-
cially available 2,6-diacetylpyridine and benzaldehydes in 1:2 stoichiometry. At 2:1 ratios, four related 6,6′′-diacety-
lated-4′-arylterpyridines were isolated in 70–73% yields in one-pot condensations in the presence of NH3. 4,4′-
Azobenzaldehyde, prepared from nitrobenzaldehyde in three steps and 40% overall yield was similarly converted to a
novel azo-linked bis(terpyridine) in 50% yield in a reaction that assembles seven molecules in one step. The 6,6′′-
diacetylated-4′-arylterpyridines and the correspondingly substituted 2,6-dicinnamoylpyridines were condensed in 1:1 ra-
tio together with NH3 to form 4,4′′,4IV-triarylcyclosexipyridines in 22–26% yields. These were obtained as mixed Na+
and K+ complexes and were insoluble amorphous solids, except for one example bearing 4-neopentoxyphenyl substitu-
1
ents. H NMR showed that the 4,4′′,4IV-tri(4-neopentoxyphenyl)cyclosexipyridine complexes form aggregates in solu-
tion and at low concentrations show twofold symmetry arising from a loss of planarity.
Key words: terpyridines, 4′-aryl-6,6′′-diacetylterpyridines, azo-bisterpyridine, cyclosexipyridines, macrocyclization.
Résumé : La condensation de la 2,6-diacétylpyridine avec des benzaldéhydes commercialement disponibles, à une stoe-
chiométrie de1:2, a permis d’obtenir quatre 2,6-dicinnamoylpyridines avec des rendements de 60 à 65 %. Lors de
condensations monotopes, à des rapports de 2:1, en présence d’ammoniac, les quatre 6,6′′-diacétyl-4′-arylterpyridines
apparentées peuvent être isolées avec des rendements de 70 à 73 %. Le 4,4′-azobenzaldéhyde qui a été préparé en trois
étapes et avec un rendement global de 40 % à partir du nitrobenzaldéhyde a été converti, avec un rendement de 50 %,
en une nouvelle bis(terpyridine) liée par une fonction azo qui permet d’assembler sept molécules en une seule étape.
Les 6,6′′-diacétyl-4′-arylterpyridines et les 2,6-dicinnamoylpyridines substituées de la même façon peuvent être conden-
sées l’une à l’autre dans un rapport de 1:1 en présence d’ammoniac, pour former de 4,4′′,4IV-triarylcyclosexipyridines
avec des rendements allant de 22 à 26 %. Celles-ci sont obtenues sous la forme de complexes avec des ions Na+ et K+
et il s’agit de solides amorphes insolubles, à l’exception de celle portant des subtituants 4-néopentoxyphényles. Les spec-
1
tres RMN du H montrent que les complexes de 4,4′′,4IV-tri(4-néopentoxyphényl)cyclosexipyridines forment des agrégats
en solution et que, à faibles concentrations, ils présentent une symétrie binaire qui découle d’une perte de planéité.
Mots clés : terpyridines, 4′-aryl-6,6′′-diacétylterpyridines, azo-bisterpyridine, cyclosexipyridines, macrocyclisation.
[Traduit par la Rédaction] Masciello and Potvin 218
Introduction
impurity in a reagent. Toner (4) reported C2-symmetrical
cyclosexipyridines (e.g., 4) prepared from bipyridines as in-
soluble and only partly characterized Na+ complexes, in
which the Na+ was apparently abstracted from the glassware.
The parent, C6-symmetric cyclosexipyridine 1 (Scheme 1)
was first reported by Newkome and Lee in 1983 (1) as the
insoluble and uncharacterized product of a seven-step syn-
thesis. Kelly et al. (2) prepared the curious, inverted cyclo-
sexipyridine 2, a highly insoluble and partly characterized
C3-symmetric tris(bipyridine), after a lengthy synthetic se-
quence. Bell and Firestone (3) used Friedlander reactions to
assemble torand 3, a C3-symmetric hexaazakekulene deriva-
tive with rigidifying ethylene bridges and solubilizing butyl
side-chains, isolated in 3% yield after 11 steps as a Ca2+
complex in which the Ca2+ apparently originated as a minute
Macrocyclic polypyridines such as cyclosexipyridine con-
stitute interesting, rigid supramolecular hosts for cations and
H-bond donors, and this was the motivation for the earlier
work. With competition experiments, Bell et al. showed that
the related torands, with stronger imino dipoles and a more
rigid architecture than in crown ethers, are extremely strong
complexing agents (5). We anticipated that cyclosexipyri-
dines would also be interesting ligands for transition metals
in that they would provide a planar, conjugated π system oc-
cupying the equatorial plane of the metal, as do the porphy-
rins and phthalocyanines. However, unlike those ligands, the
cyclopolypyridines are neutral and do not form a contiguous
aromatic system. The macroligand LUMO would thus re-
main relatively high energy, as would any MLCT transition
and photochemistry therefrom, and the reduction potentials
would remain relatively more negative. The binding geome-
Received 31 July 2002. Published on the NRC Research
Press Web site at http://canjchem.nrc.ca on 19 March 2003.
L. Masciello and P.G. Potvin.1 Department of Chemistry,
York University, 4700 Keele Street, Toronto, ON M3J 1P3,
Canada.
1Corresponding author (e-mail: pgpotvin@yorku.ca).
Can. J. Chem. 81: 209–218 (2003)
doi: 10.1139/V03-020
© 2003 NRC Canada

210
Can. J. Chem. Vol. 81, 2003
Scheme 1.
try could be trigonal or (distorted) square planar and the
binding may be fluxional.
ments and % relative intensities. UV–visible spectra were
obtained at room temperature on a Hewlett-Packard 8452A
diode array spectrophotometer. IR spectra were obtained on
a Genesis II FT-IR spectrometer. Elemental analyses were
preformed by Guelph Chemical Laboratories Ltd. in Guelph,
Ont.
DMF and CH2Cl2 were made anhydrous by distillation of
reagent grade material over MgSO4 or P2O5, respectively,
and stored over 4 Å molecular sieves. All other materials
were used as received, except for p-tolualdehyde and 4-tert-
butylbenzaldehyde, which were washed with 5% aq. NaOH,
dried (MgSO4), and then stored under Ar. Thin layer chro-
matography used Macherey–Nagel Alugram silica gel plates.
Preparative thin layer chromatography used EM Science sil-
ica gel 60F glass plates. The petroleum ether (PE) used had
a bp range of 60–80°C.
For studies of cyclosexipyridine coordination, one would
need a short, high-yield synthesis of material. In the course
of our work with Ru(II) complexes of nonmacrocyclic poly-
pyridines, we developed a one-step route from commercially
available materials to symmetrical dipyrazinylpyridines (6)
and recently applied it to the synthesis of two phenylene-
linked bisterpyridines (7, 8). This provided an opening to an
expedient synthesis of novel C3-symmetric cyclosexipyri-
dines (e.g., 5), which, we reasoned, should be more soluble
than Toner’s C2-symmetric varieties yet much easier to pre-
pare than were the torands. Furthermore, our ability to pre-
pare bisterpyridines provides an entry to an unprecedented
bis(cyclosexipyridine). In this paper, we report the prepara-
tion of a number of diacetylated terpyridines and an azo-
linked bisterpyridine that are of distinct interest and on
macrocyclization reactions to form C3-symmetric
cyclosexipyridines.
2,6-Di-(3-p-tolylpropenoyl)pyridine (7a)
A solution of 2,6-diacetylpyridine (3.08 g, 1.9 mmol) in
MeOH (40 mL) was added to a solution of p-tolualdehyde
(4.99 g, 42 mmol) in a 1:2 mixture of MeOH and 5% aq.
KOH (60 mL), and the resulting mixture was stirred vigor-
ously at room temperature for 48 h. The precipitate was iso-
lated by vacuum filtration. A second crop was obtained by
adding 30 mL of H2O to the filtrate, allowing it to stir for
8 h, and collecting the precipitate by vacuum filtration. The
combined crops were dissolved in CHCl3 and extracted with
H2O (3 × 25 mL), 5% aq. NaHSO3 (2 × 25 mL), and brine
(2 × 25 mL). The combined organic fractions were dried
with MgSO4, filtered, and freed of solvent in vacuo. The
Experimental section
General
NMR spectra were acquired on a Bruker AMX 400 MHz
spectrometer in CDCl3 unless otherwise indicated. A Kratos
Profile mass spectrometer was used for EI-MS acquisitions.
MALDI-MS spectra were acquired on a Voyager-DE STR
instrument. High resolution mass spectroscopy was per-
formed by the McMaster Regional Centre for Mass Spec-
trometry in Hamilton, Ont. Mass peaks are presented as
mass-to-charge ratios (m/z) with, in brackets, peak assign-
© 2003 NRC Canada

Masciello and Potvin
211
resulting yellow residue was recrystallized (CHCl3:MeOH)
to afford yellow needles (4.51 g, 65%). mp 198–200°C. H
by 1H NMR) by preparative TLC (5:95 EtOAc:PE). The ma-
jor UV-active band was scraped off and extracted by
sonicating in 30:70 MeOH:CH2Cl2. The extract was filtered,
reduced in vacuo, and dried to afford a colourless oil
(0.70 g, 42%). The 1H NMR, EI-MS, IR, and elemental
analysis data were in agreement with the literature data (10).
1
NMR (ppm) δ: 8.43 (d, J = 16 Hz, 2H, H-3′), 8.39 (d, J =
7.5 Hz, 2H, H-3), 8.10 (t, J = 7.9 Hz, 1H, H-4), 8.03 (d, J =
16 Hz, 2H, H-2′), 7.69 (d, J = 8.0 Hz, 4H, Ar H-3), 7.29 (d,
J = 7.1 Hz, 4H, Ar H-2), 2.46 (s, 6H, CH3). 13C NMR (ppm)
δ: 188.7, 153.3, 145.1, 141.4, 138.3, 132.5, 129.8, 128.8,
125.8, 119.8, 21.7; EI-MS: 367 ([M+], 100). Elemental anal-
ysis calcd. for C25H21NO2: C 81.72, H 5.76, N 3.81; found:
C 81.79, H 5.36, N 3.89.
2,6-Di-(3-(4-neopentoxyphenyl)propenoyl)pyridine (7d)
By analogy to the preparation of 7a, a mixture of 2,6-
diacetylpyridine (56.2 mg, 0.345 mmol) in MeOH (5 mL)
and 4-neopentoxybenzaldehyde (130.5 mg, 0.680 mmol) re-
acted at room temperature for 48 h. As with 7a, working up
provided a yellow residue that recrystallized (MeOH:H2O)
as a bright yellow powder (111 mg, 63%). mp 143–145°C.
1H NMR (ppm) δ: 8.39 (d, J = 7.9 Hz, 2H, H-3), 8.35 (d, J =
16 Hz, 2H, H-3′), 8.08 (t, J = 7.9 Hz, 1H, H-4),7.9 (d, J =
16 Hz, 2H, H-2′), 7.73 (d, J = 8.3 Hz, 4H, Ar H-3), 6.70 (d,
J = 8.5 Hz, 4H, Ar H-2), 3.70 (s, 4H, CH2), 1.09 (s, 18H,
CH3). EI-MS: 511 ([M+], 100). Elemental analysis calcd. for
C33H37NO4: C 77.47, H 7.29, N 2.74; found: C 77.21, H
7.52, N 2.73.
2,6-Di-(3-(4-chlorophenyl)propenoyl)pyridine (7b)
A mixture of 2,6-diacetylpyridine (1.28 g, 7.82 mmol) in
MeOH (20 mL) and 4-chlorobenzaldehyde (2.39 g,
17.2 mmol) in 1:2 MeOH:15% aq. KOH (60 mL) reacted at
room temperature for 2 d and the product mixture was
worked up as with 7a to provide a yellow residue that
recrystallized (cold CHCl3) as a pale yellow powder (1.97 g,
1
62%). mp 210–212°C. H NMR (warm CDCl3) (ppm) δ:
8.41 (d, J = 8.8 Hz, 2H, H-3), 8.38 (d, J = 16 Hz, 2H, H-3),
8.12 (t, J = 8.6 Hz, 1H, H-4), 7.99 (d, J = 16 Hz, 2H, H-2′),
7.69 (d, J = 9.1 Hz, 4H, Ar H-3), 7.45 (d, J = 8.9 Hz, 4H, Ar
H-2). EI-MS: 407 ([M+], 100). Elemental analysis calcd. for
C23H15NO2Cl2: C 67.66, H 3.70, N 3.43; found: C 67.37, H
3.62, N 3.44.
3-(p-Tolyl)-1-(2-(6-acetylpyridyl))-2-propenone (8a)
A solution of 2,6-diacetylpyridine (3.08 g, 18.9 mmol) in
MeOH (50 mL) was added to a vigorously stirring solution
of p-tolualdehyde (2.23 g, 18.6 mmol) in a 1:4 mixture
(50 mL) of MeOH and 2.5% aq. KOH with a syringe pump
at a rate of one drop per 10 s. The resulting mixture was
stirred vigorously at room temperature for 24 h. The precipi-
tate was isolated by vacuum filtration. A second crop was
obtained by adding 30 mL of H2O to the filtrate, allowing it
to stir for 8 h, and collecting the precipitate by vacuum fil-
tration. The combined crops were dissolved in CHCl3 and
extracted with H2O (3 × 25 mL), 5% aq. NaHSO3 (2 ×
25 mL), and brine (2 × 25 mL). The organic fraction was
dried with MgSO4, filtered, and freed of solvent in vacuo.
The residue was sonicated in MeOH to precipitate the con-
taminating bispropenone 7a as a yellow powder (0.58 g)
from the mixture. The yellow filtrate was freed of solvent in
vacuo and the remaining yellow residue recrystallized
(MeOH:H2O) as yellow flakes (3.31 g, 67%). mp 114–
2,6-Di-(3-(4-tert-butylphenyl)propenoyl)pyridine (7c)
By an entirely analogous process, 2,6-diacetylpyridine
(1.02 g, 6.25 mmol) in MeOH (25 mL) and 4-tert-butyl-
benzaldehyde (2.51 g, 15.5 mmol) in 1:2 MeOH:5% aq.
KOH (75 mL) reacted for 2 d at room temperature, and the
product mixture was worked up as with 7a to provide a yel-
low residue that was converted to a yellow powder by
sonication in MeOH (25 mL) and then recrystallized
(CHCl3:MeOH) to afford yellow needles (1.69 g, 60%). mp
1
169–171°C. H NMR (ppm) δ: 8.4 (d, J = 16 Hz, 2H, H-3′),
8.40 (d, J = 7.8 Hz, 2H, H-3), 8.10 (t, J = 7.7 Hz, 1H, H-4),
8.05 (d, J = 16 Hz, 2H, H-2′), 7.76 (d, J = 8.5 Hz, 4H, Ar H-
3), 7.52 (d, J = 8.3 Hz, 4H, Ar H-2), 1.40 (s, 18H). 13C
NMR (ppm) δ: 188.5, 154.5, 153.2, 144.9, 138.4, 132.6,
128.6, 126.0, 125.7, 120.2, 35.0, 31.2. EI-MS: 451 ([M+],
100). Elemental analysis calcd. for C31H33NO2: C 82.44, H
7.37, N 3.10; found: C 82.48, H 7.81, N 3.16.
1
116°C. H NMR (ppm) δ: 8.37 (d, J = 7.7 Hz, 1H, H-3/5),
8.30 (d, J = 16 Hz, 1H, H-3′), 8.24 (d, J = 7.8 Hz, 1H, H-
5/3), 8.04 (t, J = 7.8 Hz, 1H, H-4), 7.98 (d, J = 16 Hz, 1H,
H-2′), 7.63 (d, J = 8.1 Hz, 2H, Ar H-3), 7.28 (d, J = 8.0 Hz,
2H, Ar H-2), 2.89 (s, 3H, COCH3), 2.44 (s, 3H, Ar CH3).
13C NMR (ppm) δ: 199.4, 188.6, 153.7, 152.6, 145.7, 141.4,
137.9, 132.7, 129.3, 128.8, 126.1, 124.5, 119.4, 25.8, 22.2.
EI-MS: 265 ([M+], 100). Elemental analysis calcd. for
C17H15NO2: C 76.96, H 5.70, N 5.28; found: C 76.55, H
5.51, N 5.27.
4-Neopentoxybenzaldehyde
A suspension of NaH (1.25 g, 52.1 mmol) in anhydrous
DMF (15 mL) was vigorously stirred in a three-neck round-
bottom flask fitted with a condenser and an Ar inlet. A
dropwise solution of 4-hydroxybenzaldehyde (1.05 g,
8.59 mmol) in anhydrous DMF (4 mL) was added to this
suspension. After 10 min, a solution of neopentyl tosylate
(9) (3.93 g, 16.4 mmol) and tetrabutylammonium iodide
(0.34 g) in anhydrous DMF (5 mL) was added by syringe
pump at a rate of 1 drop per 10 s. The solution was heated to
reflux for 2 d, poured into H2O (100 mL), and extracted with
diethyl ether (4 × 100 mL). The combined ether layers were
concentrated and then washed with 5% aq. NaOH (2 ×
50 mL) and brine (2 × 50 mL). The organic fraction was
dried with MgSO4, filtered, and freed of solvent in vacuo.
The residue was dissolved in a small amount of CHCl3 and
separated from contaminants (starting materials, identified
Alternatively, the residue containing the two products can
be separated by column chromatography (silica, gradient
eluent beginning with 3:95 EtOAc:PE and adding 5:95
EtOAc:PE) to afford each as pure yellow products.
3-(4-tert-Butylphenyl)-1-(2-(6-acetylpyridyl))-2-
propenone (8c)
As for 8a, 2,6-diacetylpyridine (1.25 g, 7.7 mmol) in
MeOH (30 mL) and p-tert-butylbenzaldehyde (1.25 g,
7.7 mmol) in 1:2 MeOH:2.5% aq. KOH (90 mL) reacted for
© 2003 NRC Canada

212
Can. J. Chem. Vol. 81, 2003
24 h, and the reaction mixture was worked up as before. The
residue was dissolved in a minimal amount of EtOAc and
passed down a short column (silica, 10:90 EtOAc:PE). The
major yellow fore-fraction was collected, freed of solvent in
vacuo, and sonicated in MeOH (25 mL) to precipitate the
bispropenone 7c in pure form (0.18 g). The filtrate was freed
of solvent and the yellow residue recrystallized
(MeOH:H2O) as yellow needles (0.44 g, 19%). mp 83–85°C.
1H NMR (ppm) δ: 8.39 (d, J = 7.8 Hz, 1H, H-3/5), 8.35 (d,
J = 16 Hz, 1H, H-3′), 8.26 (d, J = 7.7 Hz, 1H, H-5/3), 8.05
(t, J = 8.0 Hz, 1H, H-4), 8.02 (d, J = 16 Hz, 1H, H-2′), 7.70
(d, J = 8.4 Hz, 2H, Ar H-3), 7.51 (d, J = 8.4 Hz, 2H, Ar H-
2), 2.90 (s, 3H, COCH3), 1.38 (s, 9H, C(CH3)3). EI-MS: 307
([M+], 100). Elemental analysis calcd. for C20H21NO2: C
78.15, H 6.88, N 4.55; found: C 78.29, H 7.00, N 4.49.
acted with 4-chlorobenzaldehyde (819 mg, 5.85 mmol) in
MeOH (20 mL), 15% aq. KOH (7 mL) and concd. NH4OH
(10 mL) at room temperature for 3 days, and the reaction
mixture was worked up as before. The yellow solid crude
product was passed down a short column (silica, 20:80
EtOAc:PE) with slow elution of the major yellow band and
10:90 MeOH:CH2Cl2 for elution of the baseline fraction.
Both fractions were combined and recrystallized
(CHCl3:MeOH) to afford a pure brown-yellow powder
1
(1.80 g, 72%). mp 166–168°C. H NMR (ppm) δ: 8.89 (d,
J = 7.8 Hz, 2H, H-5), 8.82 (s, 2H, H-3′), 8.15 (d, J = 7.4 Hz,
2H, H-3), 8.06 (t, J = 7.8 Hz, 2H, H-4), 7.83 (d, J = 8.5 Hz,
2H, Ar H-3), 7.57 (d, J = 8.41 Hz, 2H, Ar H-2), 2.91 (s, 6H,
CH3). EI-MS: 427 ([M+], 100). Elemental analysis calcd. for
C25H18N3O2Cl: C 70.18, H 4.24, N 9.82; found: C 69.70, H
4.24, N 9.57.
6,6′′-Diacetyl-4′-(p-tolyl)-2,2′:6′,2′′-terpyridine (9a)
6,6′′-Diacetyl-4′-(4-tert-butylphenyl)-2,2′:6′,2′′-
terpyridine (9c)
One-pot method A
A solution of 2,6-diacetylpyridine (1.51 g, 9.27 mmol) in
MeOH (35 mL) was quickly added to a vigorously stirring
solution of p-tolualdehyde (0.56 g, 4.64 mmol) in MeOH
(10 mL), 30% aq. KOH (7.5 mL), and concd. NH4OH
(7.5 mL). An additional 30 mL of NH4OH was added and the
mixture was allowed to stir at room temperature for 4–5 d.
The precipitate was filtered by vacuum filtration, washed
with H2O, and dissolved in CHCl3. The solution was washed
sequentially with H2O (2 × 25 mL) and brine (2 × 25 mL).
The organic fraction was dried with MgSO4 and freed of sol-
vent in vacuo. The residue was sonicated in MeOH (30 mL)
to afford a golden yellow powder that was filtered and
recrystallized (CHCl3:MeOH) as a brown-yellow powder
By analogy to the preparation of 9a by Method A, 2,6-
diacetylpyridine (1.503 g, 9.22 mmol) in MeOH (35 mL) re-
acted with 4-tert-butylbenzaldehyde (0.742 g, 4.58 mmol) in
MeOH (10 mL), 30% aq. KOH (7.5 mL) and concd. NH4OH
(7.5 mL) at room temperature for 3–4 days, and the reaction
mixture was worked up as before. The yellow solid crude
product was dissolved in a small amount of CHCl3 and sepa-
rated from contaminants by preparative TLC (silica, 1:9
EtOAc:hexane). The major UV-active band was extracted
from the silica with 30:70 MeOH:CH2Cl2, filtered, and re-
duced in vacuo to afford a white residue. Sonicating in
MeOH (30 min,100 mL) and recrystallization (CHCl3:Et2O)
afforded a white powder (1.51 g, 73%). mp 185–187°C
(dec.). 1H NMR (ppm) δ: 8.88 (d, J = 7.8 Hz, 2H, H-5), 8.85
(s, 2H, H-3), 8.12 (d, J = 7.6 Hz, 2H, H-3), 8.04 (t, J =
7.7 Hz, 2H, H-4), 7.86 (d, J = 8.1 Hz, 2H, Ar H-3), 7.64 (d,
J = 8.2 Hz, 2H, Ar H-2), 2.93 (s, 6H, COCH3), 1.74 (s, 9H,
C(CH3)3). EI-MS: 449 ([M+], 100). Elemental analysis
calcd. for C29H27N3O2: C 77.48, H 6.05, N 9.35; found: C
77.82, H 6.34, N 8.89.
1
(0.76 g, 70%). mp 191–193°C (dec.). H NMR (ppm) δ:
8.89 (d, J = 7.6 Hz, 2H, H-5), 8.85 (s, 2H, H-3′), 8.14 (d, J =
7.7 Hz, 2H, H-3), 8.03 (t, J = 7.7 Hz, 2H, H-4), 7.81 (d, J =
7.8 Hz, 2H, Ar H-3), 7.43 (d, J = 8.0, 2H, Ar H-2), 2.91 (s,
6H, COCH3), 2.50 (s, 3H, Ar CH3). EI-MS: 407 ([M+], 68).
365 ([M+-COCH3], 45), 43 ([COCH3], 100). Elemental anal-
ysis calcd. for C26H21N3O2: C 76.64, H 5.19, N 10.31;
found: C 76.26, H 5.05, N 10.65.
6,6′′-Diacetyl-4′-(4-neopentoxyphenyl)-2,2′:6′,2′′-
terpyridine (9d)
Two-step method B
A solution of 2,6-diacetylpyridine (409 mg, 2.51 mmol) in
MeOH (20 mL) was quickly added to a vigorously stirring
solution of the propenone 8a (664 mg, 2.51 mmol) in MeOH
(62.5 mL) and 15% aq. KOH (22 mL). Concentrated
NH4OH (60 mL) was added in portions and the mixture was
allowed to stir at room temperature for 4–5 d. The precipi-
tate was filtered by vacuum filtration, dissolved in CHCl3,
and washed sequentially with H2O (2 × 25 mL) and brine
(2 × 25 mL). The organic fraction was dried with MgSO4
and freed of solvent in vacuo. The residue was triturated
with MeOH (20 mL) to afford a golden yellow powder that
was filtered and recrystallized (CHCl3:MeOH) as a brown-
By analogy to the preparation of 9a by Method A, 2,6-
diacetylpyridine (335 mg, 2.05 mmol) in MeOH (5 mL) re-
acted with 4-neopentoxybenzaldehyde (191 mg, 0.99 mmol)
in MeOH (10 mL), 30% aq. KOH (2.5 mL), and concd.
NH4OH (2.5 mL) at room temperature for 4 days, and the
reaction mixture was worked up as before. Purification of
the yellow solid crude product by preparative TLC (1:3
EtOAc:hexane) gave a major yellow band that was extracted
from the silica with 30:70 MeOH:CH2Cl2, filtered, and re-
duced in vacuo. The yellow solid proved difficult to
recrystallize and was isolated as a golden-yellow powder by
1
drying under vacuum (334 mg, 70%). mp 167–170°C. H
1
yellow powder (358 mg, 35%). The mp and H NMR data
were identical to those from material obtained by Method A.
NMR (ppm) δ: 8.88 (d, J = 7.8 Hz, 2H, H-5), 8.83 (s, 2H,
H-3′), 8.13 (d, J = 7.7 Hz, 2H, H-3), 8.05 (t, J = 7.6 Hz, 2H,
H-4), 7.85 (d, J = 8.2 Hz, 2H, Ar H-3), 7.15 (d, J = 8.1 Hz,
2H, Ar H-2), 3.75 (s, 2H, CH2), 2.92 (s, 6H, COCH3), 1.10
(s, 9H, C(CH3)3). EI-MS: 479 ([M+], 87). Elemental analysis
calcd. for C30H29N3O3: C 75.16, H 6.10, N 8.79; found: C
74.80, 6.49, 8.48.
6,6′′-Diacetyl-4′-(4-chlorophenyl)-2,2′:6′,2′′-terpyridine
(9b)
By analogy to the preparation of 9a by Method A, 2,6-
diacetylpyridine (1.960 g, 12.0 mmol) in MeOH (40 mL) re-
© 2003 NRC Canada

Masciello and Potvin
213
4′,4′′′,4v-Tri-(p-tolyl)-[(2,6)-pyridyl-6-corand-6] (5a)
3/5E, H-2A or H-2/6E), 3.75 (6H, s, CH2), 2.89 (3H, s, OAc
CH3), 1.13 (27H, s, CH3). MALDI-MS: 971.2 ([M+ + Na+],
100). 987.4 ([M+ + K+], 37).
Method A
A mixture of 9a (40.1 mg, 0.098 mmol) and bispropenone
7a (37.9 mg, 0.103 mmol) were heated at reflux in glacial
acetic acid (25 mL) and excess concd. NH4OAc (1.3 g) for
24 h. The precipitate was filtered hot and washed with CHCl3,
MeOH, and Et2O to give a dark black solid (16.5 mg, 22%).
This solid was found to be insoluble in all common organic
solvents examined. mp 207–210°C (dec.). MALDI-MS: 755.1
([M+ + Na+], 100). 771.0 ([M+ + K+], 7). HRMS calcd. for
C51H36N6K: 771.264; found: 771.262 0.005.
4′-(p-Tolyl)-6,6′′-di-(3-(p-tolyl)propenoyl)-2,2′:6′,2′′-
terpyridine (10)
A solution of 9a (100.3 mg, 0.25 mmol) in THF (20 mL)
was added, dropwise, to a vigorously stirring solution of p-
tolualdehyde (208.4 mg, 1.74 mmol), MeOH (23 mL), and
15% aq. KOH (25 mL), and the mixture was stirred for 3 d.
The precipitate was filtered, dissolved in CHCl3, and washed
sequentially with 5% aq. NaHSO3, (2 × 15 mL), H2O (2 ×
15 mL), and brine (2 × 15 mL). The product was contami-
nated with a small quantity of unreacted 9a that was par-
tially removed by seven cycles of preparative TLC (silica,
10:90 EtOAc:PE). The baseline fraction was removed from
the silica (30:70 MeOH:CH2Cl2), freed of solvent, and
recrystallized (CHCl3:MeOH) to afford a tan powder
Method B
A mixture of bispropenone 10 (5.0 mg, 8.17 µmol) and
2,6-di[2-(N-pyridyl)-1-oxoethyl]pyridine diiodide 11 (11)
(4.6 mg, 8.23 µmol) were heated to reflux in glacial acetic
acid (25 mL) containing NH4OAc (0.72 g) for 24 h. The pre-
cipitate was filtered hot and washed with CHCl3, MeOH,
and Et2O to give a dark black solid (1.23 mg, 20%), identi-
cal in all respects to material obtained by Method A.
1
(30.1 mg, 20%). H NMR (ppm) δ: 8.89 (d, J = 8.0 Hz, 2H,
H-5), 8.85 (s, 2H, H-3′), 8.53 (d, J = 16 Hz, 2H, H-3), 8.14
(d, J = 7.2 Hz, 2H, H-3), 8.05 (t, J = 7.6 Hz, 2H, H-4), 8.04
(d, J = 16 Hz, 2H, H-2), 7.81 (d, J = 8.0 Hz, 2H, Ar H-3),
7.70 (d, J = 7.9 Hz, 2H, Ar H-3), 7.43 (d, J = 7.9 Hz, 6H, Ar
H-2), 2.50, (s, 6H, Ar CH3), 2.46 (s, 3H, Ar CH3). MALDI-
MS: 611 ([M+], 87).
4′,4′′,4v-Tri-(4-chlorophenyl)-[(2,6)-pyridyl-6-corand-6]
(5b)
The preparation of 5a by Method A was followed by us-
ing 9b (43.4 mg, 0.102 mmol) and bispropenone 7b
(38.5 mg, 0.0946 mmol) to obtain a dark brown, insoluble
solid (16.7 mg, 22%). mp 198–200°C (dec.). MALDI-MS:
815.2 ([M+ + Na+], 100). 831.2 ([M+ + K+], 7). HRMS
calcd. for C48H27N6Cl3Na: 815.125; found: 815.126 0.005.
2-(4-Nitrophenyl)-1,3-dioxolane (12)
A mixture of 4-nitrobenzaldehyde (1.508 g, 9.9 mmol),
ethylene glycol (1.21 mL, 21.7 mmol), and chlorotri-
methylsilane (5.61 mL, 44.2 mmol) in anhydrous CH2Cl2
(55 mL) was stirred at reflux under Ar for 48 h. A 5% aq.
solution of NaHCO3 (100 mL) was added and the mixture
was extracted with Et2O (4 × 75 mL). The ether layer was
extracted with brine (4 × 75 mL), dried with MgSO4, fil-
tered, and reduced in vacuo. The yellow residue was
4′,4′′,4v-Tri-(4-tert-butylphenyl)-[(2,6)-pyridyl-6-corand-6]
(5c)
A solution of 9c (24.3 mg, 0.054 mmol) in 4:1 n-
butanol:MeOH (25 mL) was added to a stirring solution of
bispropenone 7c (17.0 mg, 0.0377 mmol) in 2:1 n-
butanol:MeOH (15 mL), 30% aq. KOH (10 mL) and concd.
NH4OH (30 mL). The mixture was heated at reflux for 24 h.
The precipitate was filtered hot and washed with H2O,
CHCl3, and MeOH to give a tan solid (9.0 mg, 26%) that
was insoluble in all common organic solvents examined. mp
204–207°C (dec.). MALDI-MS: 881.4 ([M+ + Na+], 25),
897.4 ([M+ + K+], 37). HRMS calcd. for C60H54N6Na and
C60H54N6K: 881.431 and 897.405; found: 881.427 0.005
and 897.408 0.005.
1
recrystallized (CHCl3) as cream flakes (1.523 g, 79%). H
NMR, EI-MS, and mp data were in agreement with those re-
ported in the literature (12, 13).
4,4′-Di-(1,3-dioxolanyl)azobenzene (13)
A suspension of LiAlH4 (1.510 g, 3.98 mmol) in Et2O
(50 mL) was vigorously stirred at –78°C in a three-neck
round-bottom flask fitted with a mechanical stirrer, dropping
funnel, and an Ar inlet. To this suspension, a solution of 2-
(4-nitrophenyl)-1,3-dioxolane 12 (0.6954 g, 3.58 mmol) in
Et2O (40 mL) was added, dropwise. The mixture was al-
lowed to slowly warm to room temperature (3 h) and stirred
for an additional 30 min. The reaction was quenched slowly
and successively with EtOAc, MeOH, and H2O with vigor-
ous stirring. The addition of a 5% aq. H2SO4 solution pro-
duced two brightly coloured (orange and yellow) layers. The
orange layer was extracted into Et2O (40 mL), and the ether
layer was washed with H2O (1 × 25 mL), NaHCO3 (2 ×
25 mL), and brine (2 × 25 mL). The ether-borne extract was
concentrated and purified by preparative TLC (silica, 1:3
EtOAc:hexane). The major orange band was extracted from
the silica with 30:70 MeOH:CH2Cl2. The extract was fil-
tered, concentrated in vacuo, and recrystallized (cold
MeOH) to afford a bright orange powder (0.7236 g, 62%).
4′,4′′,4v-Tri-(4-neopentoxyphenyl)-[(2,6)-pyridyl-6-
corand-6] (5d)
In analogy to the preparation of 5a by Method A, 9d
(42.3 mg, 0.0885 mmol) and bispropenone 7d (45.4 mg,
0.0888 mmol) were used to provide a fine precipitate, which
was filtered hot, dissolved in CHCl3, and washed with H2O
(3 × 15 mL). The organic layer was freed of solvent in vacuo
and dried under vacuum to give a yellow solid (18.5 mg,
22%). mp 173–175°C (dec.). UV–vis (CHCl3) λmax (nm) (ε
(103 M–1 cm–1)): 264 (45), 316 (49). 1H NMR (ppm)
(0.64 µM, refer to Fig. 1 for assignments) δ: 9.04 (2H, s, H-
3B), 8.95 (2H, d, H-3/5C), 8.83 (2H, s, H-3/5D), 8.83 (1H, t,
H-4F), 8.76 (2H, d, H-5/3C), 8.14 (2H, d, H-3F), 8.07 (2H,
t, H-4C), 7.97 (2H, d, H-2A or H-2/6E), 7.92 (2H, d, H-2A
or H-2/6E), 7.92 (2H, s, H-5/3D), 7.14 (8H, d, H-3A, H-
1
mp 140–142°C. H NMR (ppm) δ: 7.96 (d, J = 7.9 Hz, 4H,
H-2), 7.66 (d, J = 8.0 Hz, 4H, H-3), 5.92 (s, 2H, Ar-CH),
© 2003 NRC Canada

214
Can. J. Chem. Vol. 81, 2003
Fig. 1. Cyclosexipyridine 5d, showing twisting of rings A and F
out of plane. The H NMR assignments used the position num-
Results and discussion
1
As illustrated in Scheme 2, we explored two approaches
to the synthesis of C3-symmetrical cyclosexipyridines:
(i) coupling a diacetylterpyridine 9 with a bispropenone 7 in
the presence of a source of NH3 via a dihydropyridine inter-
mediate and (ii) coupling a more elaborate bispropenone 10
with the salt 11 under Kröhnke’s conditions (17). Both ap-
proaches required diacetylated terpyridines 9 whose synthe-
sis was to make use of our one-pot preparation of
dipyrazinylpyridines (6), essentially a modified Chichibabin
synthesis (18). This would start from commercial 2,6-
diacetylpyridine 6, which would be the only source of pre-
manufactured pyridine rings and would require a differential
reactivity of the two acetyl groups.
bers given.
The substituents on the macroring, originating from aro-
matic aldehydes, could presumably be varied at will. In prin-
ciple, this approach can also lead to less symmetrical
products with one macroring substituent different from the
other two, but this possibility was not explored here. Our
initial goal was to prepare three examples of cyclosexipy-
ridines, one with electron-releasing substituents, another
with electron-withdrawing substitution, and the first example
of a bis(cyclosexipyridine) with an azo linkage. The latter
necessitated the preparation of a novel azo-linked bisterpy-
ridine.
Diacetylpyridine 6 was treated with 2 equiv of p-tolualde-
hyde under basic conditions to afford a 65% isolated yield of
recrystallized bispropenone 7a (Scheme 2). Similarly, 4-
chlorobenzaldehyde afforded 7b in 62% isolated yield. Their
NMR and EI-MS spectra and their elemental analyses were
consistent with the proposed structures. According to 1H
NMR, they appeared to be exclusively trans,trans isomers.
When only 1 equiv of p-tolualdehyde and weaker base were
used, a 67% isolated yield of monopropenone 8a was iso-
lated, again as the trans isomer according to 1H NMR, along
with some 7a. Hence, the two acetyl groups of 6 seem to re-
act fairly independently, and the reaction of one apparently
reduces the reactivity of the other, thereby making our ap-
proach promising.
4.20–4.08 (m, 8H, CH2). EI-MS: 326 ([M+], 48). 149 ([M+ –
C9H9O2], 100). Elemental analysis calcd. for C18H18N2O4: C
66.25, H 5.56, N 8.58; found: C 66.77, H 5.69, N 8.31.
4,4′-Azobenzaldehyde (14)
A 10% aq. H2SO4 solution (100 mL) was added to a vig-
orously stirring solution of 4,4′-di-(1,3-dioxolanyl)azoben-
zene 13 (0.700 g, 2.15 mmol) in MeOH (60 mL) at room
temperature. The mixture was allowed to stir for 20 h, fil-
tered, dissolved in CHCl3, and washed sequentially with 5%
aq. NaHCO3 (1 × 75 mL), H2O (1 × 75 mL), and brine (2 ×
50 mL). The organic fraction was dried with MgSO4, freed
of solvent in vacuo, and recrystallized (MeOH) to afford red
1
leaflets (0.420 g, 82%). H NMR, EI-MS, and mp data were
in agreement with those reported in the literature (14–16).
UV–vis (CHCl3) λmax (nm) (ε (M–1 cm–1)): 336 (53 500),
478 (1 240).
4,4′-Di-(4′-(6,6′′-diacetyl-2,2′:6′,2′′-terpyridinyl))azoben-
zene (15)
A solution of 2,6-diacetylpyridine (0.4890 g, 3.0 mmol) in
MeOH (10 mL) was added slowly to a vigorously stirring
mixture of 4,4′-azobenzaldehyde 14 (0.1017 g, 0.49 mmol),
MeOH (90 mL), 15% aq. KOH (2.5 mL), and concd.
NH4OH (23 mL). The mixture was heated to reflux for 20 h.
The orange precipitate was filtered and washed with H2O.
Trituration with CHCl3 afforded an orange filtrate and an in-
soluble orange precipitate. The filtrate was washed with
NaHCO3 (2 × 25 mL), NaOH (1 × 25 mL), and brine (2 ×
25 mL), dried with MgSO4, and freed of solvent in vacuo.
The residue was purified by preparative TLC (silica, 31:2:1
MeCN: saturated KNO3:H2O). The baseline fraction was ex-
tracted from the silica with 30:70 MeOH:CH2Cl2. The ex-
tract was filtered and concentrated in vacuo. The orange
solid recrystallized (CHCl3:MeOH) as an orange powder
(175.0 mg, 50%). mp 207–209°C. UV–vis (CHCl3) λmax
(nm) (ε (M–1 cm–1)): 356 (12 400), 274 (31 400). IR (neat)
(cm–1): 1698 (s, C=O). 1H NMR (ppm) δ: 8.94 (d, J =
7.6 Hz, 4H, H-5), 8.91 (s, 4H, H-3′), 8.23 (d, J = 7.6 Hz, 4H,
H-3), 8.15 (d+d, J = 7.8 Hz, 8H, Ar H-2 & H-3), 8.10 (t, J =
7.5 Hz, 4H, H-4), 2.94 (s, 12H, CH3). MALDI-MS: 812
([M+], 100). HRMS calcd. for C50H37N8O4: 813.294. found:
813.295 0.005.
When 6 and propenone 8a were combined in 1:1 ratio un-
der basic conditions in the presence of NH3 at room temper-
ature, a precipitate formed over the course of several days.
After collection, washing, and recrystallization, a 35% yield
of diacetylterpyridine 9a was obtained. If, instead, p-
tolualdehyde or 4-chlorobenzaldehyde were combined with
2 equiv of 6 in the presence of NH3, terpyridines 9a and 9b
precipitated in essentially pure form over a few days at room
temperature. These terpyridines were isolated in 70% and
72% yields, respectively, after recrystallization. Given the
number of intermediates involved in the assembly of the
central pyridine ring, such yields are remarkable. Both of
these terpyridines were fully characterized by NMR, EI-MS,
and elemental analyses. A distinctive feature of these materi-
1
als was the appearance of a new aromatic singlet in their H
NMR spectra, signalling the creation of the symmetrical,
central pyridine ring.
By an analogous process, 4,4′-azobenzaldehyde (14) and
4 equiv of 6 were converted to the novel azo-linked
bis(terpyridine) 15, an orange solid, in 50% yield after pre-
parative TLC and recrystallization (Scheme 3). Considering
© 2003 NRC Canada

Masciello and Potvin
215
Scheme 2. Conditions: (a) ≥2 equiv Ar-CHO, 5–15% aq. KOH, MeOH, 2 d; (b) 1 equiv Ar-CHO, 2.5% aq. KOH, MeOH, 24 h; (c) 1
equiv 6, 15% aq. KOH, concd. NH4OH, MeOH, 4–5 d; (d) 0.5 equiv 6, 15–30% aq. KOH, concd. NH4OH, MeOH, 3–5 d; (e) 1 equiv
7a,b, or d, NH4OAc, HOAc, reflux, 24 h, or 1 equiv 7c, 30% aq. KOH, concd. NH4OH, MeOH:n-BuOH, reflux, 24 h; (f) 7 equiv
1
CH3C6H4-4-CHO, 15% aq. KOH, MeOH, 3 d; (g) 1 equiv 11, NH4OAc, HOAc, reflux, 24 h. The H NMR assignments used the posi-
tion numbers given.
that this transformation assembled a total of seven molecules
(including NH3) and eliminated a total of six molecules of
H2O (as well as 2 equiv of H2), the 50% yield is quite re-
markable. For this synthesis, the known dialdehyde 14 (14–
16) was prepared from 4-nitrobenzaldehyde involving its
TMSCl-catalyzed (19) protection to nitroacetal 12 (12, 13)
(79% yield), reductive coupling with cold LiAlH4 (20) to the
previously unknown azo-acetal intermediate, the bright or-
ange 13 (62% yield), followed by deprotection (82% yield).
The nitro–azo transformation caused an upfield migration of
the aromatic signals and the structure of 13 was otherwise
supported by MS and elemental analysis as well as its trans-
formation to 14. In the 1H NMR spectrum of 15, the
azophenyl doublets (identified by COSY) were overlapping.
This had been also observed with the aromatic doublets
from 14 where the aldehyde group is conjugated with the
azo group but not in 12 or 13 where the aldehyde is masked.
The overlap with 15 is therefore consistent with conjugation
of the terpyridine unit with the azo group and hence π-over-
lap among the terpyridines. The high degree of conjugation
present was also evident from the energy of the π→π* band
for 15 (356 nm), which is at a lower energy than had been
found with a phenylene-linked bis(terpyridine) (280 nm) (7)
or with azobenzene itself (320 nm) (21, 22).
These terpyridine compounds are also of evident interest
in transition metal complexation. The synthesis provides for
variation of the 4′ substituent, and the products may be fur-
ther elaborated at the acetyl groups, for instance, to prepare
macrocyclic Schiff bases with diamines.
To effect macrocyclization, 7a and 9a were combined in
1:1 ratio and heated to reflux in acetic acid in the presence
of excess NH4OAc. This produced a black powder in 22%
yield that was insoluble in all solvents tested and appeared
to decompose on the melting point apparatus. NMR analysis
© 2003 NRC Canada

216
Can. J. Chem. Vol. 81, 2003
Scheme 3. Conditions: (a) LiAlH4, Et2O, –78°C → RT, 30 min;
(b) 10% aq. H2SO4, MeOH, 20 h; (c) 6 equiv 6, 15% aq. KOH,
concd. NH4OH, MeOH, reflux, 20 h. The H NMR assignments
material prepared earlier and in a similar yield (20%). There
was, therefore, no apparent advantage to applying the
Krönhke method and aerial oxidation was evidently not a
limiting factor. We also attempted the direct, one-pot ap-
proach that was successful in our terpyridine synthesis using
a 1:1 ratio of 6 and p-tolualdehyde in the presence of NH3.
However, the major product was the monopropenone 8a. It
appears that, once the first acetyl group of 6 has reacted to
give 8a, the reaction of the second is too slow in Michael
addition to the propenone moiety of 8a.
1
used the position numbers given.
By analogy to other flat, extended π-conjugated systems
such as phthalocyanines (23, 24), the insolubility of these
cyclosexipyridines was likely due to aggregation. To dis-
courage this, we turned to the commercially available 4-tert-
butylbenzaldehyde to deliver bulkier substituents. We were
able to prepare the tert-butylated bispropenone 7c and
terpyridine 9c in yields very similar to those obtained previ-
ously (60 and 73%, respectively), but the monopropenone 8c
was obtained in just 19% yield, in part because of difficulty
encountered in separating it from the 7c that accompanies it,
and 8c was not used further. The macrocyclization was per-
formed this time under basic conditions analogous to those
used in our terpyridine synthesis, as opposed to the acidic
conditions used earlier for 5a and 5b, to control the cation
content of the reaction mixture. Thus, terpyridine 9c and
bispropenone 7c were treated with aqueous KOH and con-
centrated NH4OH in a butanol–MeOH mixture to produce a
tan deposit in 26% yield that again analyzed as a mix of Na+
and K+ complexes of cyclosexipyridine 5c, this time in a K-
rich ratio of ca. 2:3 Na:K, with OH– as the likely counterion.
The Na+ probably again originated from the glass reaction
vessel, and the failure of the large excess of K+ in solution to
prevent the incorporation of Na+ suggests that the latter
forms the stronger complexes. Indeed, Bell et al. measured
equilibrium binding constants for torand 2 and found a ca.
2.5-fold preference for Na+ over K+ (5). This contrasts with
the usual preference for K+ by crowns (25) and discounts the
primacy of cavity size in cation selection, i.e., the “hole-size
concept”, by which the cyclosexipyridine cavity size, esti-
mated at 1.3 Å (26), better fits the hexacoordinate ionic ra-
dius of K+ (1.38 Å) than that of Na+ (1.02 Å) and is
therefore expected to better bind K+.
of the mother liquor was not informative. MALDI-MS anal-
ysis of the insoluble material revealed two principal signals
consistent with the presence of both Na+ (M + 23) and K+
(M + 39) complexes of cyclosexipyridine 5a, in roughly
15:1 ratio, with no sign of metal-free macrocycle. This result
is consistent with Toner’s obtention of his cyclosexipyridine
as a NaOAc complex. It seems that the presence of a third
substituent on our macroring failed to produce a more solu-
ble product. Essentially the same result was obtained by the
analogous combination of the chlorophenyl-bearing
terpyridine 9b with chlorophenyl–bispropenone 7b; a 22%
yield of an insoluble black powder analyzing as Na+ and K+
complexes of macrocycle 5b in ca. 15:1 ratio.
The condensation reactions forming the last two pyridine
rings of the macrocycle actually produce dihydropyridines
that require aerial oxidation. In case this might be rate-
limiting and contribute to the low yields of macrocycles 5,
an alternative synthesis was performed using Krönhke meth-
odology (17) such as was applied by Toner (4) and by Kelly
et al. (2). This leads to fully formed pyridines without need
for aerial oxidation but requires extra steps in substrate prep-
aration. Diacetylterpyridine 9a was converted to the
trans,trans-bispropenone 10 by base-catalyzed condensation
with excess p-tolualdehyde. This gave a disappointing,
though unoptimized, yield of 20% of material that was,
moreover, contaminated with some unreacted 9a that proved
to be very tedious to remove. Compound 10 was not fully
characterized, but its structure was proven by MALDI-MS
In any case, macrocycle 5c was just as insoluble as the pre-
vious two examples. We then turned to neopentoxy groups to
provide solubilizing power, as had been used with success to
prepare soluble phthalocyanines (27). There is a two-step
published preparation of 4-neopentoxybenzaldehyde (10), but
we prepared it by alkylation of 4-hydroxybenzaldehyde using
neopentyl tosylate (9) in 42% isolated yield. Conversions to
bispropenone 7d and terpyridine 9d proceeded in the same
fashion and in similar yields (63% 7d, 70% 9d) as for their
analogues. Also, the macrocyclization proceeded as well as
the previous examples (22% isolated yield) but this time the
precipitated product was soluble in a variety of solvents,
though not aromatic ones. This material defied all attempts at
crystallization. TLC with various eluents showed that it was
chromatographically homogeneous and free of impurities.
MALDI-MS identified the material as 5d, again obtained as a
mix of Na+ and K+ complexes in ca. 5:2 ratio.
1
and H NMR, which showed the absence of acetyl CH3 sig-
nals, the appearance of trans-alkenyl signals, and the pres-
ence of two benzylic CH3 signals in 2:1 integration ratio.
Bispropenone 10 was subsequently treated with the known
dipyridinium salt 11 and excess NH4OAc in acetic acid at re-
flux, according to the Krönhke method (17). This afforded
the Na+ and K+ complexes of 5a identical in all respects to
1H NMR showed broad aliphatic singlets for this material,
one for C(CH3)3 and one for OCH2 in 9:2 ratio, in accord
© 2003 NRC Canada

Masciello and Potvin
217
with the structure, as well as one attributable to the OAc
counterion, but the aromatic region contained broad and
overlapping signals, indicating a pronounced tendency to ag-
gregate in spite of being well dissolved. Heating the sample
to 62°C failed to affect the spectrum, but dilution caused a
sharpening of the signals. Progressively sharper signals were
obtained with progressive dilution, confirming that 5d, as a
Na+:K+ complex, was subject to aggregation. Even at its
sharpest, at sub-micromolar concentrations, the NMR spec-
trum was not readily interpretable, and it became evident
that the symmetry was lower than threefold. Since COSY
could not be performed at such high dilutions, homo-
decoupling experiments were carried out to help assign the
signals, and a total of 11 aromatic signals were thereby lo-
cated. Their number, multiplicities, and integration values
were consistent with twofold symmetry, allowing for some
accidental overlap, compared with the five signals expected
with the threefold symmetry. Twofold symmetry may result
from an out-of-plane twist of any one of the six pyridine
rings or of any transannularly opposed pair of rings, as de-
picted in Fig. 1. Such out-of-plane twisting was indeed pres-
ent in the crystal structures of the torand alkali metal
complexes (26) and was predicted by computation to occur
with the Na+ and Li+ complexes of the parent cyclosexi-
pyridine (28), but those structures were of D3-symmetry,
with the pyridine rings twisting “up” then “down” in alterna-
tion, a reduction from the sixfold symmetry of the flat lig-
ands. The D3 conformation probably minimizes lone pair –
lone pair repulsions and contacts between pyridine H-3/5.
The same conformation with the less symmetrical com-
plexes of 5d would place all three neopentoxy groups on the
same face of the macroring, which may be destabilizing.
Twofold symmetry would give rise to a maximum 14 aro-
matic signals, 12 with free rotation of the side-chains. Our
analysis of the 11 signals observed implies that one of the
H-2/6 2H doublets has suffered an upfield shift of about
0.8 ppm, consistent with an out-of-plane twist of one
phenylene ring such that its H-2 and H-6 are thereby less
deshielded by the neighbouring pyridines. This analysis also
implies that the two equivalent side-chains (rings E), if not
all three, are not freely rotating owing to conjugation and
(or) aggregation.
be insoluble in all solvents tested, as was true with 5a–c.
MALDI-MS revealed numerous high-mass species, but none
at or near m/z 1798, the molecular weight expected for the
desired bis(cyclosexipyridine). No further cyclization at-
tempts were made.
Conclusion
We have extended our one-pot synthesis of dipyrazinyl-
pyridines and phenylene-linked bisterpyridines to four exam-
ples of terpyridines and one example of an azo-linked
bisterpyridine. The simple terpyridines were converted to
novel C3-symmetrical triarylated cyclosexipyridines without
resorting to the Kröhnke method, but this was only soluble
with neopentoxy substitution. Aggregation and demetalation
remain issues to be resolved.
Acknowledgement
The authors thank the Natural Sciences and Engineering
Research Council of Canada (NSERC) for funding.
References
1. G.R. Newkome and H.W. Lee. J. Am. Chem. Soc. 105, 5956
(1983).
2. T.R. Kelly, Y.J. Lee, and R.J. Mears. J. Org. Chem. 62, 2774
(1997).
3. T.W. Bell and A. Firestone. J. Am. Chem. Soc. 108, 8109
(1986).
4. J.L. Toner. Tetrahedron Lett. 24, 2707 (1983).
5. T.W. Bell, A. Firestone, and R. Ludwig. J. Chem. Soc. Chem.
Commun. 1902 (1989).
6. R. Liegghio, P.G. Potvin, and A.B.P. Lever. Inorg. Chem. 40,
5485 (2001).
7. S. Vaduvescu. M.Sc. thesis, York University, Toronto, Ont.
2002.
8. S. Vaduvescu and P.G. Potvin. Inorg. Chem. 41, 4081 (2002).
9. F.M. Beringer and H.S. Schultz. J. Am. Chem. Soc. 77, 5533
(1955).
10. R.A. Widenhoefer and S.L. Buchwald. J. Am. Chem. Soc. 120,
6504 (1998).
Metal-free torand 3 was isolated after macrocyclization
catalyzed by triflic acid (3). We attempted to prepare 5d in
the same manner from 9d and 7d using excess NH4OAc in
glacial acetic acid spiked with 4% triflic acid. However,
heating this to reflux for 3 d failed to cause any precipita-
tion, and NMR indicated the presence of only the starting
materials. Triflic acid apparently inhibited the macro-
cyclization, likely because of its stronger acidity and the fa-
vourable protonation of the pyridine rings in the starting
materials.
Interestingly, sonicating a CDCl3 solution of 5d with a so-
lution of potassium picrate in D2O for 24 h failed to cause
any detectable anion or cation exchange according to NMR
and MALDI-MS. Encapsulating hosts also showed slow ion
exchanges (5).
11. L.C. King. J. Am. Chem. Soc. 66, 894 (1944).
12. H.E. Baumgarten, D.L. Pederson, and M.W. Hunt. J. Am.
Chem. Soc. 80, 1977 (1958).
13. W.C. Danen, C.T. West, T.T. Kensler, and T.J. Tipton. J. Am.
Chem. Soc. 94, 4830 (1972).
14. F.J. Always and W.D. Bonner. J. Am. Chem. Soc. 27, 1107
(1905).
15. P. Freundler. Comptes Rendus, 134, 1360 (1905).
16. J.M. Khurana and A. Ray. Bull. Chem. Soc. Jpn. 69, 407
(1996).
17. F. Kröhnke. Synthesis, 1 (1976).
18. A.E. Chichibabin. J. Russ. Phys. Chem. Soc. 37, 1229 (1906).
19. T.H. Chan, M.A. Brook, and T. Chaly. Synthesis, 203 (1983).
20. R.F. Nystrom and W.G. Brown. J. Am. Chem. Soc. 70, 3738
(1948).
21. H. Suzuki. Electronic absorption spectra and geometry of or-
ganic molecules. Academic Press, New York. 1967. Chap 23.
22. P.P. Birnbaum, J.H. Linford, and W.G. Doyle. Trans. Faraday
Soc. 49, 735 (1953).
Azo-bis(terpyridine) 15 was also subjected to our
macrocyclization conditions with 2.2 equiv of 7d for 48 h
and a precipitate formed as before. Though four neopentoxy
groups were intended to be present, this material proved to
23. M. Hanack and M. Lang. Adv. Mater. 6, 822 (1994).
© 2003 NRC Canada

218
Can. J. Chem. Vol. 81, 2003
24. E. Sielcken, M. Van Tilborg, M.F.M. Roks, W. Drenth, and
J.M. Nolte. J. Am. Chem. Soc. 109, 4261 (1987).
25. P.G Potvin and J.M. Lehn. In Synthesis of macrocycles: The
design of selective complexing agents. Edited by R.M. Izatt
and J.J. Christensen. John Wiley & Sons, New York. 1987.
p. 167.
26. T.W. Bell, P.J. Cragg, M.G. Drew, and A.M. Firestone. Angew.
Chem. Int. Ed. 31, 345 (1992).
27. C.C. Leznoff, S.M. Marcuccio, S. Greenberg, A.B.P. Lever,
and K.B. Tomer. Can. J. Chem. 63, 623 (1985).
28. S.T. Howard and I.A. Fallis. J. Chem. Soc., Perkin Trans. 2,
2501 (1999).
© 2003 NRC Canada